Summary In April 1978, Chevron Canada Resources Ltd. initiated cyclic steam injection into a 21.5-acre [87 000-m2] seven-spot pilot on Cold Lake Lease No. 79. Steam was injected above fracture pressure to achieve the desired injection rates. Early observation well temperature responses and later pressure, temperature, and production responses in the producing wells indicated a production responses in the producing wells indicated a northeast-southwest fracture orientation. Heat movement away from this northeast-southwest trend was relatively slow except in the vicinity of one well. Simultaneous steaming of on-trend wells appeared to accelerate the off-trend heat movement and apparently changed the direction of the fracture orientation. Significant drive response was observed between one on-trend well pair. The cumulative steam/oil ratio (SOR) for completed cycles was about six in Nov. 1981. Oil recovery was about 6.3% of the original oil within the pilot drainage area. A fractured reservoir steam-injection simulator was used to study cyclic steam and steam drive responses. Simulation of cyclic steaming of a single well showed thatincreasing the steam slug size improves oil recovery for slug sizes ranging from 7.5 to 30.0 × 10(3) stock-tank bbls [1192 to 4770 stock-tank m3], andincreasing the slug size from cycle to cycle gives better oil recovery than a fixed slug size. A northeast-southwest fracture orientation was assumed in the simulation of pilot cyclic operation followed by steam drive. The resulting heat communication between on-trend wells caused depletion of the mobile oil along the trend and little incremental oil production during steam drive. Heat movement off-trend was very slow because of mobility barriers created by banking of cold oil and the development of irreducible water saturations. By reducing the spacing between off-trend wells, heat communication was achieved and mobility barriers were avoided. Introduction Field tests of steam injection to recover extremely viscous, heavy oil from the Cold Lake deposits of Alberta began in 1964. Esso Resources Canada Ltd. has operated a succession of steam-injection pilots in the Clearwater formation. On the basis of their pilot results, they designed a commercial-scale project to recover 160,000 B/D [25 440 m3/d] of bitumen. This project was suspended indefinitely in 1981 but Esso still project was suspended indefinitely in 1981 but Esso still produces oil from two heavy-oil projects. Eleven produces oil from two heavy-oil projects. Eleven companies are currently operating 16 experimental projects in the Cold Lake heavy-oil deposits. In 1975 Chevron Canada Resources Ltd. conducted a single-well, cyclic steam pilot in the B channel sand of the Lower Grand Rapids formation (Fig. 1). Excellent production response (180 B/D [28.6 m3/d] of oil for 7 production response (180 B/D [28.6 m3/d] of oil for 7 days) was obtained during the first cycle until well sanding and mechanical failure caused a shutdown. During steam injection in the second cycle, steam broke through to the Lower B sand, which had bottom water at that location. Squeeze cementing was unsuccessful in shutting off the water. The resulting high water and low oil production caused termination of the project after the production caused termination of the project after the third cycle. Encouraged by the early results of the single-well pilot, Chevron initiated a seven-well steam-injection pilot, Chevron initiated a seven-well steam-injection pilot in 1976. The general objective of the pilot was to pilot in 1976. The general objective of the pilot was to evaluate the producing potential of the B channel sand and other sands of the Grand Rapids formation. Specific objectives are described in Ref. 6.To simulate the cyclic steam response of the Cold Lake heavy oil sands mathematically, a fractured tar sand simulator was developed at Chevron Oil Field Research Co. This simulator was used to evaluate the effect of steam slug size on cyclic response from a single well and the effect of converting to steam drive on pilot response. This paper reports on performance of the pilot and the results of the simulation study. Geology A brief description of the geology is helpful in understanding the pilot performance and the design of the simulation study. The Cold Lake bitumen deposits contain an estimated 164 × 10(9)bbls [26.1 × 10 (9) m3] of bitumen in place. Chevron's Lease No. 79, site of the steam-injection pilot, contains an estimated 1.1 × 10(9) bbls [0.175 × 10(9) pilot, contains an estimated 1.1 × 10(9) bbls [0.175 × 10(9) m3] of bitumen distributed among sands in the Upper and Lower Grand Rapids formations. The relationship of Upper and Lower Grand Rapids sands and shales is presented schematically in Fig. 1. presented schematically in Fig. 1. The target sand for the steam-injection pilot is the B channel sand in the Lower Grand Rapids formation. This sand trends southwest to northeast, which could have some bearing on the preferred direction of heat movement during pilot steam injection. It averages 35 ft [10.7 m] thick and occurs at a depth of about 1,200 ft [366 m]. JPT P. 1781